Hardwired GC problem

[Sherman]

This past weekend I had the opportunity to show off my GC built with Brian's boards to a friend. He was blown away by the overall simplicity of it as well as the sound.

When he got home he immediately began building one with a chip I gave him. He inserted the components through perf board and soldered hookup wire under the board to put it together. He said he used a picture of one of Brian's PCBs to do the layout.

However when he hooks it up with input from a CD player and powers it up all he gets is a very loud buzz from the speaker (it is only one channel as I only had one extra chip to give him). I haven't heard it yet myself but he will bring it over tomorrow.

As he was using materials at hand his resistors are Dale 1% and the caps are cheap 1000uF Xicons. He wired all the input/output lines to a Radio Shack PC mount terminal block (except the CHG line since the terminal block only has 8 positions). His volume pot is a 100K audio taper (not stereo).

The ps caps are 47uF. The ps is using a 44VCT toroid and it is giving him +/-31VDC so all seems well there.

He emailed a sketch to me today and I cleaned it up a bit. It is attached. The arrangement looks OK to me.

Any ideas? I'd like to help him get it working when he brings it over tomorrow. BTW I have strongly suggested he buy one of Brian's kits!

[officeboy]

Missing feedback, Something like this should work a "bit" better.

[officeboy]

Oh and Tie all those grounds together too. Forgot to fix that. Maybe use a shared Signal ground and Power Ground and then jumper from one to the other.

[Sherman]

The NFB resistor is there. (It isn't on the drawing, but is noted at the bottom of it.) It is soldered directly to pins 3 and 8.

As for grounds- I think you are on the right track, a ground loop or something but it appears to me that the grounds are tied together though not as "cleanly" as they could be.

PG+ and OG are different ends of the same wire which loops around and is soldered to C1's negative lead. Similarly, PG- and CHG are ends of the same wire soldered to C2's + lead. He also put in a jumper from the PG+/OG side to the PG-/CHG side.

I had him short the input and give it a quick test with the same result. He doesn't want to leave it on for more than a second for fear that he will damage some components so he hasn't measured DC offset yet either.

[officeboy]

It's SG and PG that need to be connected.

[Sherman]

Originally posted by officeboy
It's SG and PG that need to be connected.

Officeboy,
Thanks for the help! I gave my buddy a call and suggested that he check all solder joints again, especially the Rnfb and the grounds. I then suggested that he tie the signal ground to PG- since those points are next to each other on his setup.

Bingo! There is music coming from his speaker! Now he has to go out and buy more LM 3875s and some decent capacitors (though he says the Xicons don't sound all that bad but I'll reserve judgement until I can here it tomorrow).

Thanks again,